Norris Boyd (NS) Andrew Clinick (MS) Mike Cowlishaw (IBM, Project Editor) Brendan Eich (NS) Waldemar Horwat (NS) Roger Lawrence (NS) Clayton Lewis (NS) Drew Little (MS) Karl Matzke (SunSoft) Mike McCabe (NS) Herman Venter (MS) Rok Yu (MS)
The morning was spent reviewing the expected comments on the ISO DIS from the USA national body, confirming their correct resolution in the 1998.01.27 draft for Version 1 (V1), and agreeing minor corrections to that draft. The editor will apply the changes to both versions of the documents; Karl Matzke took notes and will prepare the disposition of comments.
The following lists were propagated from the notes of the 1997.10.10 meeting and updated:
caller (omitted from V1) do while break to label continue to label switch regexp === operator (strict equality) conditional compilation literal notation function closures (expression, nesting) reveal __parent__ , __photo__ arguments object exception handling toSource (people want a way to make objects persistent) Function.prototype.apply instanceof
binary object Date (as presented by Borland in 1997) generic sequence operations on a string or an array threading issues undefined literal, not reserved parse {int, float} step point result toString extensions date to string toBoolean (object) Hide proto.property meta object protocol (MOP) package concept
The remainder of the meeting discussed various items from the first list in more detail.
Herman Venter (MS) described his proposal
NS proposal described by Waldemar Horwat
After some discussion the committee agreed on:
Issues:
Array literal constructors, proposed:
o=new Array(0,1,2) b=new Array(0) a=[0,1,2] b=[0] c=[0,,2] d=[0,1,,] -- d.length = 3 e=[,1] -- e.length = 2
Object literal constructors
o=new Object o.p=1 -- same as o['p']=1 o['*hi*']=2 o[2]=true o={p:1, '*hi*':2, 2:true} -- ok o2={null:3, true:4, false:5} -- ok, but no o3={(x+1):6} -- badIssues:
x5=#1={self:#1#}
Under discussion; some members of the committee did not feel that this proposal added much value.
(The draft does not use sharp variables yet.) Generally agreed. If this becomes a built-in language item, the proposal may not require the use of sharp variables.